Skip to content
Help centre

WordPress updates

Update plugins, themes and WordPress safely, one at a time with a backup, automatically in a maintenance window or by hand, with staging, visual checks and failure tickets.

Once a WordPress site has the OpsDash Connector plugin, OpsDash can update its plugins, themes and WordPress itself. Each update run backs the site up, applies updates one at a time, checks the site after each one, and rolls an update back if it breaks the site.

The WordPress updates page

Site health › WordPress updates lists your connected WordPress sites with their pending updates, connector version and latest run. Open a site to see:

  • the connector, WordPress and PHP versions, and how many updates are pending,
  • the maintenance window (when automatic updates run), and its staging copy,
  • Packages: every plugin and theme with its installed and available versions and its policy,
  • Runs: every update run, newest first.

Check now asks the site to check in and report its versions.

Package policies

Each plugin or theme has a policy:

  • Auto: updated in the maintenance window.
  • Manual: only when someone chooses it.
  • Blocked: never updated (for example a plugin you've customised).

New packages get the site's default: auto-update in the window, wait for a person, or never. WordPress core is off until you turn it on.

Running updates

  • Update selected updates the packages you tick.
  • Update Auto updates every package set to Auto.
  • The maintenance window (a day, hour and timezone) runs automatically each week, updating every Auto package, including major and premium updates the site knows about.

Choose Stop the run when an update fails if you'd rather skip the rest after a rollback; by default the run carries on with the others as long as the site is healthy again.

Staging first

If a production site has a staging site connected (set on the production site's settings), maintenance windows run on staging first, then the same updates are copied to production. OpsDash doesn't create staging servers: connect your staging copy after its connector has checked in.

Visual check after updates

In Settings › Site health › Visual check after updates, turn on Take before and after screenshots (needs a ScreenshotOne key; off by default). The homepage and the site's PageSpeed pages are screenshotted before and after each run. If a page changed more than the threshold (2% of its pixels catches a broken layout without flagging small text changes), the run is marked Needs review, and a staging run waits for someone to choose Looks fine, approve before it's copied to production. Compare Side by side, or see Changes with changed pixels in red.

What a run's status means

StatusMeaning
Waiting for the siteQueued; OpsDash is waiting for the site to check in and start.
RunningUpdating, one package at a time.
UpdatedEvery package updated and the site is healthy.
Finished with problemsIt went through every package, but some weren't updated (rolled back or couldn't install).
StoppedIt stopped part-way and skipped the rest.
Rolled backThe update broke the site, so the backup was restored.
Needs reviewThe visual check found a page that changed a lot.
Failed / CancelledIt couldn't run, or someone cancelled it.

Open a run to see each package's step, what was checked and by whom, PHP errors before and after, and what WordPress said. See what the run messages mean.

When an update fails

Owners and admins, and anyone chosen to be alerted on the site, hear about it. In Settings › Site health › WordPress updates you can also open a ticket when a WordPress update fails (off until switched on): one ticket for the site's client, saying which package, what went wrong and whether the backup was restored. The client isn't emailed; PHP error details go in an internal note.

The emergency stop

Pause all WordPress updates and restores, in Site health settings, is an emergency stop: while it's on, no update run starts (by hand or in a window), and OpsDash sends your sites no updates, backups or restores. Runs already going stop where they are. Switch it off to carry on.

Updates can also be held on a site, for example while its database is being repaired, or if its connector is too old to back up safely. The site's page says why.

Tidying up

Delete a finished run, or Delete older runs, to tidy the page. Backups and restore history are kept.

These open in OpsDash. You’ll be asked to sign in first if you aren’t already.